Trust Region-Guided Proximal Policy Optimization

Source code for the paper: Trust Region-Guided Proximal Policy Optmization (TRGPPO). The original code was forked from OpenAI baselines.

Method is tested on MuJoCo continuous control tasks and Atari discrete game tasks in OpenAI gym. Networks are trained using tensorflow1.10 and Python 3.6.

Installation

git clone --recursive https://github.com/wangyuhuix/TRGPPO
cd TRGPPO
pip install -r requirements.txt

Usage

Command Line arguments

• env: environment ID
• seed: random seed
• num_timesteps: number of timesteps

Continuous Task

python -m baselines.ppo2_AdaClip.run --env=InvertedPendulum-v2 --seed=0

Discrete Task

python -m baselines.ppo2_AdaClip.run --env=BeamRiderNoFrameskip-v4 --seed=0 --isatari